Bloomberg,
Cicada, Deutsche Bank, Fidelity Investments, FISD, FT Interactive,
Goldman Sachs, ILX, London Market Systems, London Stock Exchange,
Merrill Lynch, Moneyline Telerate, Reuters, Thomson Financial,
Telekurs Financial
Purpose of Meeting
The objective of this meeting was to design a process for defining
the market data vocabulary for the debt domain. Once the Vocabulary
Committee defines the universe of terms, the Technical Committee will
be able to create a MDDL structure. This activity is critical in that
vocabulary definition is the process that drives MDDL. The functional
coverage matrix (described below) should be viewed as a tool/mechanism
to allow the industry to define the universe of terms needed to
specify fixed income instruments – regardless of the level of
instrument complexity.
Executive Summary
- Members specified the methodology for defining the vocabulary
structure of the debt instrument domain using the functional
coverage matrix developed by London Market Systems. The matrix
provides a logical framework for identifying the universe of
properties (terms) from an instrument, applications and market
segment perspective.
- FISD and SWIFT agreed (in principle) to combine MDDL into the
15022 data dictionary. The idea is to create a single repository of
market data and transactions-related terms - with normalized and
agreed upon definitions. MDDL terminology will also be mapped to ISO
standard Classification of Financial Instrument (CFI) definitions as
part of the process. FISD and the Bond Market Association (BMA) have
initiated similar discussions on the development of the vocabulary
for bonds. Both 15022 and BMA start with XML for transactions and
are moving toward security master information. MDDL starts with
pricing and security master information. The goal is to integrate
all application types into a single glossary.
- FISD members have finalized the MDDL change management approach
as outlined during the meeting on January 11. Our challenge is to
implement the procedures on the web and to create a bulletin board
notification system to keep members informed about MDDL version
enhancements and specification changes.
We recommend going back to Yahoo Groups. We searched and
failed to find a better in-house system. Is this a satisfactory
solution?
Debt Hierarchy
This
link is a visual representation outlining the scope in
instruments/asset types within MDDL. It specifies the debt domain (as
organized by FISD) into floating rate, fixed interest, index-linked,
backed and other instruments.
Functional Coverage Matrix
Most of the discussion during the working meeting was based on the
functional coverage matrix developed by
London Market Systems.
It is intended to provide a multi-dimensional view of the data model
for bonds. The (approximately 300) terms outlined on the left side of
the spreadsheet define the core properties associated with bond
information. The terms/properties have been organized into seven
primary categories:
- Birthright Information (basic characteristics needed to define
the issue)
- Market Information (data specific to the place of issuance)
- Price/Yield/Accrued Interest
- Derived Status Information
- Analytical Information (information about yields, durations,
convexity, life, etc.)
- Benchmark Information (benchmarks used and their spread)
- Agency Ratings (agency, rating, status, type)
The remaining sheets (market, instrument and functional) are used
to illustrate a business (not a data) view of bonds. The market
sheet allows the viewer to evaluate the terms from a market (i.e.
corporate, municipal, mortgage lender) perspective. The instrument
sheet indicates broad coverage and allows the viewer to evaluate
the completeness of terms from any specific instrument (i.e. vanilla,
callable, puttable, warrants, convertibles, etc.) perspective. The
functional sheet allows the viewer to evaluate terms based on how
the data is going to be used (i.e. pricing/performance, risk
management, position keeping, order management, etc.).
The key to the functional coverage matrix is flexibility. Experts
should be able to focus on a specific view of the world and use this
matrix as a mechanism to evaluate the completeness of terms. For
example, if the viewer only focuses on municipal bonds for portfolio
management/decision support, they will be able to use this matrix to
make sure that MDDL can define the union of attributes for any
instrument, market or application.
Data Model
FISD is looking at this matrix as the vehicle to promote evaluation
of the vocabulary for bonds. John Bottega (Merrill Lynch) volunteered
to use Unified Modeling Language (UML) diagrams to visually define
overall architecture of MDDL. Together these tools should provide a
basic structural diagram of MDDL to make sure it meets the defined
objectives.
ISO 15022 Data Dictionary
FISD met with SWIFT (the registration authority for ISO 15022) and
agreed on the value of merging MDDL terminology into the 15022 data
dictionary. The overall objectives are completeness and precision of
terms. The creation of a single repository should simplify the process
of normalization, maintenance and acceptance of both terminology and
definition. As part of that process, MDDL will also be mapped against
the ISO standard Classification of Financial Instruments (CFI).
Bond Market Association
The Bond Market Association (BMA) has approached FISD about what we
are doing with MDDL. Most of the BMA efforts have been focused on the
transactions side of the business. However, they have convened a new
Securities Master Database Task Force designed to address data quality
issues associated with reference data. They are looking at the
problems associated with data collection. They are exploring the
potential of a single source (portal) for issuer information – but are
not building an industry-standard security master file. Development of
the data model including standard formats and definitions are included
in the charter.
We have created a liaison group to work with the BMA. The
discussions are exploratory and the BMA has asked for a face-to-face
meeting as a follow-up to our February 22 conference call. This looks
promising.
Next Steps
- FISD to convert functional coverage matrix to HTML with
“instructions on how to use” and mechanism to suggesting new terms
- John Bottega (Merrill) to create visual data model for MDDL
- Sara Banerjee (Telekurs) to cross reference CFI with MDDL terms
- Annabel Saunders (Reuters) to lead effort to define a specific
bond type (or range of bond types) as the first functional step
forward within the debt domain.
- Mike Bennett (London Market Systems) to cross-reference the
terms used in V1.0 with the functional coverage matrix for bonds
(extend, don’t rebuild) and lead the glossary evaluation effort
(Saunders, Banerjee, Draper, Geddis, Robinson)
- FISD to solidify the working relationships with SWIFT (on
integration of MDDL into 15022) and with the Bond Marketing
Association (Banerjee, Bottega, Robinson)
